An experimental prostatitis has been successfully created in 30 dogs by injecting E. coli into a branch of the prostatic artery. Inflammation was proven histologically in all cases and by the appearance of E. coli in the prostatic secretion in all but two dogs. However, in these two dogs, cultures of prostatic tissue were positive for E. coli. Therefore, we now have a reproducible model in which we can investigate the distribution of antimicrobial substances, not only in the normal, but also in the infected prostrate. We have carried out infusion experiments with various antimicrobial substances in normal dogs and in dogs with experimental bacterial prostatitis and have determined the "prostatic pharmacokinetics" of trimethoprim, sulfamethoxazole, erythromycin, doxycycline and ampicillin. We intend to investigate some of these same antibacterial substances and others used in the treatment of bacterial prostatitis. This will be done by implanting small plastic amounts, but large enough for bioassay determination. This should further elucidate the distribution of various antimicrobial agents in the area of the prostate where the infection actually takes place. Parallel to this, we are investigating the concentration of antimicrobial agents in human prostatic tissue removed at prostatectomy.